Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 5 Sep 2001 22:13:47 +1000 (EST)
From:      Bruce Evans <bde@zeta.org.au>
To:        "Brian F. Feldman" <green@FreeBSD.ORG>
Cc:        Warner Losh <imp@harmony.village.org>, Mike Barcroft <mike@FreeBSD.ORG>, David Hill <david@phobia.ms>, <audit@FreeBSD.ORG>
Subject:   Re: users.c PATCH - please review 
Message-ID:  <20010905220327.C23159-100000@alphplex.bde.org>
In-Reply-To: <200109032120.f83LKoS53035@green.bikeshed.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 3 Sep 2001, Brian F. Feldman wrote:

> Warner Losh <imp@harmony.village.org> wrote:
> > In message <20010901125802.F85574@coffee.q9media.com> Mike Barcroft writes:
> > : David Hill <david@phobia.ms> writes:
> > : > I have patched users/users.c so that it will compile under WARNS?=2.  I have both attached and cut/pasted it to the message.
> > : > Please review it.
> > :
> > : > diff -ru /usr/src/usr.bin/users.orig/users.c /usr/src/usr.bin/users/users.c
> > : > --- /usr/src/usr.bin/users.orig/users.c Fri Aug 27 21:07:14 1999
> > : > +++ /usr/src/usr.bin/users/users.c      Fri Aug 31 15:58:20 2001
> > : > @@ -117,5 +117,8 @@
> > : >  scmp(p, q)
> > : >         const void *p, *q;
> > : >  {
> > : > -       return(strncmp((char *)p, (char *)q, UT_NAMESIZE));
> > : > +       const char *const pp1 = p;
> > : > +       const char *const pp2 = q;
> > : > +
> > : > +       return(strncmp(pp1, pp2, UT_NAMESIZE));
> > : >  }
> > :
> > : Why create new local variables?  Just get rid of the bogus casts.
> >
> > Why not just
> > 	return (strncmp((const char *)p, (const char *)q, UT_NAMESIZE));
> >
> > Also, const char *const is wrong.  No need for the second const.
>
> Technically, there's nothing wrong with declaring "a const pointer to a
> const character", just casting to that would be very strange (and probably
> illegal; not sure).  In any case, I'm not sure I see the point of not just
> using two (const char *) casts and not increasing the function size other
> than adding a blank line and space for style(9).  That should remove any
> warning about  casting away const qualification.

Why not just get rid of the casts like Mike suggested?  They have no effect
in ISO C.

Bruce


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-audit" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010905220327.C23159-100000>